Darwin, Meet Frankenstein

Scientists have figured out many ways to study the origin of species. They can build evoluitonary trees, to see how species descend from a common ancestor. They can survey islands or mountains or lakes to see how ecological conditions foster the rise of new species. They can look for fossils that offer clues to how long ago species branched off from one another, and how their ranges spread or shrank. Now comes a new trick in tomorrow's issue of the journal Nature: to test their ideas about how a new species of butterfly came to be, they essentially recreated it in their lab.

The new species in question was a graceful little butterfly called Heliconius heurippa. It lives in a small region of Columbia, just east of the Andes. Many other Heliconius (or passion-vine) butterflies live in South and Central America. Of particular interest is H. melpomene. Its wide range includes that of H. heurippa. The two butterflies have the same hindwing color, and they also share an orange streak across the forewing.

Scientists were also struck by the similarity between H. heurippa and another species, H. cydno. These two species share the same white streak across the wing. It does not live alongside H. heurippa, though. Its range stops at the western side of the Andes.

H. heurippa almost looks like a simple combination of the other two species. A team of scientists who study the butterfly wondered if that indeed was the case. Perhaps, they speculated, a population of H. cydno became isolated on the eastern side of the Andes from the other butterflies of its species. The isolated butterflies mated with the H. melpomene around them. Instead of merging with the other species, the butterflies became distinctive hybrids with an appearance all their own. They were attracted to their own appearance, while the original two species continued to be attracted to theirs. The hybrids became a new species: H. heurippa.

This was a provocative idea. Mathematical models had suggested that new species formed from hybrids should be very rare. It's hard for hybrids to become reproductively isolated from their pure-bred parents, particularly when they live side by side. As I wrote in the New York Times a couple weeks ago, interbreeding may be able to cause two newly diverged species to collapse back into a single hybrid swarm. The best example of new species from hybrids come from plants--in particular, sunflowers. The evidence from animals has been suggestive, but it couldn't rule out the possibility that another process had produced the same patterns. Instead of forming a new hybrid species, the animals might have first split into two new species, and then only interbred later, mixing some of their genes.

To try to rule out these sorts of alternative explanations, the scientists looked at H. heurippa from many angles. They compared its DNA to that of the other two species. All three butterfly species are closely related, but their genes belong to three distinct clusters. H. heurippa is not just some isolated race of either of the more widespread species.

On the other hand, the scientists could identify certain genes in H. heurippa that were most similar to the corresponding in genes in one of the two other species--just as you'd expect from hybrids.

Then the scientists did a Frankenstein-meets-Darwin experiment. They tried to recreate H. heurippa all over again. They mated H. melpomene and H. cydno, and produced hybrids. In the first generation, the female hybrids were all sterile--a common result of interbreeding. But this did not mean the experiment was over. The male hybrids could still breed with the females of the original species. After a few generations of interbreeding this way, the scientists produced butterflies that looked remarkably like H. heurippa and could now breed among themselves. While fertile hybrids are also familiar to biologists, they tend to produce a jumble of offspring, many of which look more like their pure-bred ancestors. Not so for the new butterflies. They consistently produced offspring that also looked like H. heurippa.

Even if hybrids can produce a stable line of offspring like themselves, they may not actually do so. If the hybrids are willing to mate with their parental species, they may simply merge back together. In the final part of their study, the scientists showed that this was not happening in H. heurippa. They observed how likely butterflies from each species would be to mate with other species. All three showed a strong preference for their own. The key to their preference seems to be those beautiful bands on their wings. The scientists covered over each of the bands on female butterflies to see how the new look affected how males decided to court. Removing either the red or yellow bands made male H. heurippa less likely to try to woo a female.

This research is intriguing on a lot of levels (more details appear in a paper in the Journal of Evolutionary Biology). The researchers propose a path to the origin of species from hybrids, one that the theorists haven't thought about before. And who knows how many other species might have formed that way. As I wrote here, the human and chimpanzee genomes contain hints that our ancestors may have interbred millions of years ago. Did they create some new hybrid species? It's possible that mate choice in primates is just too different from that in butterflies. We may not be so easily controlled by visual cues like colored stripes.

But I cringed when I looked at the map in the paper showing the distribution of the three species. H. heurippa exists in only in one stretch of Andean foothills and nowhere else in the world. Passion-vine butterflies can only lay their eggs on a single species of passion vine, because they have evolved defenses against that particular plant and can't feed on other passion vines. If its habitat or host plant should disappear, this wonderful natural experiment in evolution may will disappear as well.